Category Archives: Hydrogen

New hydrogen generator from REB Research

8 Replies

Here’s the new, latest version of our Me150 hydrogen generator with our wonder-secretary, Libby, shown for scale. It’s smaller and prettier than the previous version shown at left (previous version of Me150, not of secretary). Hydrogen output is 99.9999% pure, 9.5 kg/day, 75 slpm, 150 scfh H2; it generates hydrogen from methanol reforming in a membrane reactor. Pricing is $150,000. Uses about 7 gal of methanol-water ($6 worth) per kg of H2 (380 ft3). Can be used to fill weather balloons, cool electric dynamos, or provide hydrogen fuel for 2-10 fuel cell cars.

New REB Research hydrogen generator 150 scfh of 99.9999% H2 from methanol reforming

New REB Research hydrogen generator 150 scfh of 99.9999% pure H2 from methanol-water reforming against metal membranes.

Dr. Robert E. Buxbaum

Nuclear Power: the elephant of clean energy

10 Replies

As someone who heads a hydrogen energy company, REB Research, I regularly have to tip toe about nuclear power, a rather large elephant among the clean energy options. While hydrogen energy looks better than battery energy in terms of cost and energy density, neither are really energy sources; they are ways to transport energy or store it. Among non-fossil sources (sources where you don’t pollute the air massively) there is solar and wind: basically non-reliable, low density, high cost and quite polluting when you include the damage done making the devices.

Compared to these, I’m happy to report that the methanol used to make hydrogen in our membrane reactors can come from trees (anti-polluting), even tree farming isn’t all that energy dense. And then there’s uranium: plentiful, cheap and incredibly energy dense. I try to ignore how energy dense uranium is, but the cartoon below shows how hard that is to do sometimes. Nuclear power is reliable too, and energy dense; a small plant will produce between 500 and 1000 MW of power; your home uses perhaps 2 kW. You need logarithmic graph paper just to compare nuclear power to most anything else (including hydrogen):

log_scale

A tiny amount of uranium-oxide, the size of a pencil will provide as much power as hundreds of train cars full of coal. After transportation, the coal sells for about $80/ton; the sells for about $25/lb: far cheaper than the train loads of coal (there are 100-110 tons of coal to a train-car load). What’s more, while essentially all of the coal in a train car ends up in the air after it’s burnt, the waste uranium generally does not go into the air we breathe. The coal fumes are toxic, containing carcinogens, carbon monoxide, mercury, vanadium and arsenic; they are often radioactive too. All this is avoided with nuclear power unless there is a bad accident, and bad accidents are far rarer with nuclear power than, for example, with natural gas. Since Germany started shutting nuclear plants and replacing them with coal, it appears they are making all of Europe sicker).

It is true that the cost to build a nuclear plant is higher than to build a coal or gas plant, but it does not have to be: it wasn’t that way in the early days of nuclear power, nor is this true of military reactors that power our (USA) submarines and major warships. Commercial nuclear reactors cost a lot largely because of the time-cost for neighborhood approval (and they don’t always get approval). Batteries used for battery power get no safety review generally though there were two battery explosions on the Dreamliner alone, and natural gas has been known to level towns. Nuclear reactors can blow up too, as Chernobyl showed (and to a lesser extent Fukushima), but almost any design is better than Chernobyl.

The biggest worry people have with nuclear, and the biggest objection it seems to me, is escaped radiation. In a future post, I plan to go into the reality of the risk in more detail, but the worry is far worse than the reality, or far worse than the reality of other dangers (we all die of something eventually). The predicted death rate from the three-mile island accident is basically nil; Fukushima has provided little health damage (not that it’s a big comfort). Further, bizarre as this seems the thyroid cancer rate in Belarus in the wind-path of the Chernobyl plant is actually slightly lower than in the US (7 per 100,000 in Belarus compared to over 9 per 100,000 in the USA). This is clearly a statistical fluke; it’s caused, I believe, by the tendency for Russians to die of other things before they can get thyroid cancer, but it suggests that the health risks of even the worst nuclear accidents are not as bad as you might think. (BTW, Our company makes hydrogen extractors that make accidents less likely)

The biggest real radiation worry (in my opinion) is where to put the waste. Ever since President Carter closed off the option of reprocessing used fuel for re-use there has been no way to permanently get rid of waste. Further, ever since President Obama closed the Yucca Mountain burial repository there have been no satisfactory place to put the radioactive waste. Having waste sitting around above ground all over the US is a really bad option because the stuff is quite toxic. Just as the energy content of nuclear fuel is higher than most fuels, the energy content of the waste is higher. Burying it deep below a mountain in an area were no-one is likely to live seems like a good solution: sort of like putting the uranium back where it came from. And reprocessing for re-use seems like an even better solution since this gets rid of the waste permanently.

I should mention that nuclear power-derived electricity is a wonderful way to generate electricity or hydrogen for clean transportation. Further, the heat of hot springs comes from nuclear power. The healing waters that people flock to for their health is laced with isotopes (and it’s still healthy). For now, though I’ll stay in the hydrogen generator business and will ignore the clean elephant in the room. Fortunately there’s hardly any elephant poop, only lots and lots of coal and solar poop.

 

Statistics Joke

A classic statistics joke concerns a person who’s afraid to fly; he goes to a statistician who explains that planes are very, very safe, especially if you fly a respectable airline in good weather. In that case, virtually the only problem you’ll have is the possibility of a bomb on board. The fellow thinks it over and decides that flying is still too risky, so the statistician suggests he plant a bomb on the airplane, but rig it to not go off. The statistician explains: while it’s very rare to have a bomb onboard an airplane, it’s really unheard of to have two bombs on the same plane.

It’s funny because …. the statistician left out the fact that an independent variable (number of bombs) has to be truly independent. If it is independent, the likelihood is found using a poisson distribution, a non-normal distribution where the greatest likelihood is zero bombs, and there are no possibilities for a negative bomb. Poisson distributions are rarely taught in schools for some reason.

By Dr. Robert E. Buxbaum, Mar 25, 2013. If you’ve got a problem like this (particularly involving chemical engineering) you could come to my company, REB Research.

Hydrogen versus Battery Power

21 Replies

There are two major green energy choices that people are considering to power small-to-medium size, mobile applications like cars and next generation, drone airplanes: rechargeable, lithium-ion batteries and hydrogen /fuel cells. Neither choice is an energy source as such, but rather a clean energy carrier. That is, batteries and fuel cells are ways to store and concentrate energy from other sources, like solar or nuclear plants for use on the mobile platform.

Of these two, rechargeable batteries are the more familiar: they are used in computers, cell phones, automobiles, and the ill-fated, Boeing Dreamliner. Fuel cells are less familiar but not totally new: they are used to power most submarines and spy-planes, and find public use in the occasional, ‘educational’ toy. Fuel cells provided electricity for the last 30 years of space missions, and continue to power the international space station when the station is in the dark of night (about half the time). Batteries have low energy density (energy per mass or volume) but charging them is cheap and easy. Home electricity costs about 12¢/kWhr and is available in every home and shop. A cheap transformer and rectifier is all you needed to turn the alternating current electricity into DC to recharge a battery virtually anywhere. If not for the cost and weight of the batteries, the time to charge the battery (usually and hour or two), batteries would be the obvious option.

Two obvious problems with batteries are the low speed of charge and the annoyance of having to change the battery every 500 charges or so. If one runs an EV battery 3/4 of the way down and charges it every week, the battery will last 8 years. Further, battery charging takes 1-2 hours. These numbers are acceptable if you use the car only occasionally, but they get more annoying the more you use the car. By contrast, the tanks used to hold gasoline or hydrogen fill in a matter of minutes and last for decades or many thousands of fill-cycles.

Another problem with batteries is range. The weight-energy density of batteries is about 1/20 that of gasoline and about 1/10 that of hydrogen, and this affects range. While gasoline stores about 2.5 kWhr/kg including the weight of the gas tank, current Li-Ion batteries store far less than this, about 0.15 kWhr/kg. The energy density of hydrogen gas is nearly that of gasoline when the efficiency effect is included. A 100 kg of hydrogen tank at 10,000 psi will hold 8 kg of hydrogen, or enough to travel about 350 miles in a fuel-cell car. This is about as far as a gasoline car goes carrying 60 kg of tank + gasoline. This seems acceptable for long range and short-range travel, while the travel range with eVs is more limited, and will likely remain that way, see below.

The volumetric energy density of compressed hydrogen/ fuel cell systems is higher than for any battery scenario. And hydrogen tanks are far cheaper than batteries. From Battery University. http://batteryuniversity.com/learn/article/will_the_fuel_cell_have_a_second_life

The volumetric energy density of compressed hydrogen/ fuel cell systems is higher than for any battery scenario. And hydrogen tanks are far cheaper than batteries. From Battery University. http://batteryuniversity.com/learn/article/will_the_fuel_cell_have_a_second_life

Cost is perhaps the least understood problem with batteries. While electricity is cheap (cheaper than gasoline) battery power is expensive because of the high cost and limited life of batteries. Lithium-Ion batteries cost about $2000/kWhr, and give an effective 500 charge/discharge cycles; their physical life can be extended by not fully charging them, but it’s the same 500 cycles. The effective cost of the battery is thus $4/kWhr (The battery university site calculates $24/kWhr, but that seems overly pessimistic). Combined with the cost of electricity, and the losses in charging, the net cost of Li-Ion battery power is about $4.18/kWhr, several times the price of gasoline, even including the low efficiency of gasoline engines.

Hydrogen prices are much lower than battery prices, and nearly as low as gasoline, when you add in the effect of the high efficiency fuel cell engine. Hydrogen can be made on-site and compressed to 10,000 psi for less cost than gasoline, and certainly less cost than battery power. If one makes hydrogen by electrolysis of water, the cost is approximately 24¢/kWhr including the cost of the electrolysis unit.While the hydrogen tank is more expensive than a gasoline tank, it is much cheaper than a battery because the technology is simpler. Fuel cells are expensive though, and only about 50% efficient. As a result, the as-used cost of electrolysis hydrogen in a fuel cell car is about 48¢/kWhr. That’s far cheaper than battery power, but still not cheap enough to encourage the sale of FC vehicles with the current technology.

My company, REB Research provides another option for hydrogen generation: The use of a membrane reactor to make it from cheap, easy to transport liquids like methanol. Our technology can be used to make hydrogen either at the station or on-board the car. The cost of hydrogen made this way is far cheaper than from electrolysis because most of the energy comes from the methanol, and this energy is cheaper than electricity.

In our membrane reactors methanol-water (65-75% Methanol), is compressed to 350 psi, heated to 350°C, and reacted to produce hydrogen that is purified as it is made. CH3OH + H2O –> 3H2 + CO2, with the hydrogen extracted through a membrane within the reactor.

The hydrogen can be compressed to 10,000 psi and stored in a tank on board an automobile or airplane, or one can choose to run this process on-board the vehicle and generate it from liquid fuel as-needed. On-board generation provides a saving of weight, cost, and safety since you can carry methanol-water easily in a cheap tank at low pressure. The energy density of methanol-water is about 1/2 that of gasoline, but the fuel cell is about twice as efficient as a gasoline engine making the overall volumetric energy density about the same. Not including the fuel cell, the cost of energy made this way is somewhat lower than the cost of gasoline, about 25¢/kWhr; since methanol is cheaper than gasoline on a per-energy basis. Methanol is made from natural gas, coal, or trees — non-imported, low cost sources. And, best yet, trees are renewable.

Why the Boeing Dreamliner’s batteries burst into flames

14 Replies

Boeing’s Dreamliner is currently grounded due to two of their Li-Ion batteries having burst into flames, one in flight, and another on the ground. Two accidents of the same type in a small fleet is no little matter as an airplane fire can be deadly on the ground or at 50,000 feet.

The fires are particularly bad on the Dreamliner because these lithium batteries control virtually everything that goes on aboard the plane. Even without a fire, when they go out so does virtually every control and sensor. So why did they burn and what has Boeing done to take care of it? The simple reason for the fires is that management chose to use Li-Cobalt oxide batteries, the same Li-battery design that every laptop computer maker had already rejected ten years earlier when laptops using them started busting into flames. This is the battery design that caused Dell and HP to recall every computer with it. Boeing decided that they should use a massive version to control everything on their flagship airplane because it has the highest energy density see graphic below. They figured that operational management would insure safety even without the need to install any cooling or sufficient shielding.

All lithium batteries have a negative electrode (anode) that is mostly lithium. The usual chemistry is lithium metal in a graphite matrix. Lithium metal is light and readily gives off electrons; the graphite makes is somewhat less reactive. The positive electrode (cathode) is typically an oxide of some sort, and here there are options. Most current cell-phone and laptop batteries use some version of manganese nickel oxide as the anode. Lithium atoms in the anode give off electrons, become lithium ions and then travel across to the oxide making a mixed ion oxide that absorbs the electron. The process provides about 4 volts of energy differential per electron transferred. With cobalt oxide, the cathode reaction is more or less CoO2 + Li+ + e– —> LiCoO2. Sorry to say this chemistry is very unstable; the oxide itself is unstable, more unstable than MnNi or iron oxide, especially when it is fully charged, and especially when it is warm (40 degrees or warmer) 2CoO2 –> Co2O3 +1/2O2. Boeing’s safety idea was to control the charge rate in a way that overheating was not supposed to occur.

Despite the controls, it didn’t work for the two Boeing batteries that burst into flames. Perhaps it would have helped to add cooling to reduce the temperature — that’s what’s done in lap-tops and plug-in automobiles — but even with cooling the batteries might have self-destructed due to local heating effects. These batteries were massive, and there is plenty of room for one spot to get hotter than the rest; this seems to have happened in both fires, either as a cause or result. Once the cobalt oxide gets hot and oxygen is released a lithium-oxygen fire can spread to the whole battery, even if the majority is held at a low temperature. If local heating were the cause, no amount of external cooling would have helped.

battery-materials-energy-densities-battery-university

Something that would have helped was a polymer interlayer separator to keep the unstable cobalt oxide from fueling the fire; there was none. Another option is to use a more-stable cathode like iron phosphate or lithium manganese nickel. As shown in the graphic above, these stable oxides do not have the high power density of Li-cobalt oxide. When the unstable cobalt oxide decomposed there was oxygen, lithium, and heat in one space and none of the fire extinguishers on the planes could put out the fires.

The solution that Boeing has proposed and that Washington is reviewing is to leave the batteries unchanged, but to shield them in a massive titanium shield with the vapors formed on burning vented outside the airplane. The claim is that this shield will protect the passengers from the fire, if not from the loss of electricity. This does not appear to be the best solution. Airbus had planned to use the same batteries on their newest planes, but has now gone retro and plans to use Ni-Cad batteries. I don’t think that’s the best solution either. Better options, I think, are nickel metal hydride or the very stable Lithium Iron Phosphate batteries that Segway uses. Better yet would be to use fuel cells, an option that appears to be better than even the best batteries. Fuel cells are what the navy uses on submarines and what NASA uses in space. They are both more energy dense and safer than batteries. As a disclaimer, REB Research makes hydrogen generators and purifiers that are used with fuel-cell power.

More on the chemistry of Boeing’s batteries and their problems can be found on Wikipedia. You can also read an interview with the head of Tesla motors regarding his suggestions and offer of help.

 

Heisenberg joke and why water is wet

16 Replies

I love hydrogen in large part because it is a quantum fluid. To explain what that means and how that leads to water being wet, let me begin with an old quantum physics joke.

Werner Heisenberg is speeding down a highway in his car when he’s stopped by a police officer. “Do you know how fast you were going?” asks the officer. “No idea” answers Heisenberg, “but I know exactly where I am.”

The joke relates to a phenomenon of quantum physics that states that the more precisely you can know the location of something, the less precisely you can infer the speed. Thus, the fact that Heisenberg knew precisely where he was implied that he could have no idea of the car’s speed. Of course, this uncertainty is mostly seen with small things like light and electrons –and a bit with hydrogen, but hardly at all with a car or with Dr. Heisenberg himself (and that’s why it’s funny).

This funky property is related to something you may have wondered about: why is water wet? That is, why does water cling to your hands or clothes while liquid teflon repels. Even further, you may have wondered why water is a liquid at normal conditions when H2S is a gas; H2S is a heavier analog, so if one of the two were a liquid, you’d think it was H2S.

Both phenomena are understood through hydrogen behaving as the quantum car above. Oxygen atoms are pretty small, and hydrogen atoms are light enough to start behaving in a quantum way. When a hydrogen atom attaches to an oxygen atom to form part of a water molecule, its location becomes fixed rather precisely. As a result, the hydrogen atom gains velocity (the hydrogen isn’t going anywhere with this velocity, and it’s sometimes called zero-point energy), but because of this velocity or energy, its bond to the oxygen becomes looser than it would be if you had heavier hydrogen. When the oxygen of another water molecule or of a cotton cellulose molecule comes close, the hydrogen starts to hop back and forth between the two oxygen atoms. This reduces the velocity of the hydrogen atom, and stabilizes the assemblage. There is now less kinetic energy (or zero-point energy) in the system, and this stability is seen as a bond that is caused not by electron sharing but by hydrogen sharing. We call the reasonably stable bond between molecules that share a hydrogen atom this way a “hydrogen bond.” (now you know).

The hydrogen bond is why water is a liquid and is the reason water is wet. The hydrogen atom jumping between water molecules stabilizes the liquid water more than it would stabilize liquid H2S. Since sulfur atoms are bigger than oxygen atoms, the advantage of hydrogen jumping is smaller. As a result, the heat of vaporization of water is higher than that of H2S, and water is a liquid at normal conditions while H2S is a gas.

Water sticks to cotton or your skin the same way, hydrogen atoms skip between the oxygen of water molecules and of these surfaces creating a bond. It is said to whet these surfaces, and the result is that water is found to be wet. Liquid teflon does not have hydrogen atoms that can jump so there is no band that could be made from that direction (there are some hydrogen atoms on the cotton that can jump to the teflon, but there is no advantage to bonding of this sort as there are only a few hydrogen atoms, and these already jump to other oxygens in the cotton. Thus, to jump to the teflon would mean breaking a bond with other oxygen atoms in the cotton — there would be no energy advantage. This then is just one of the reasons I love hydrogen: it’s a quantum-y material.

A visit to the Buxbaum laboratory from Metromedia

3 Replies

It’s a slow news day in Detroit, so the folks from Metromedia came to visit my laboratory at REB Research. You can visit too. We’re doing cool stuff most of the time, we’re working on a hydrogen-fueled plane that stays aloft for weeks (not that cool, actually, the Hindenberg did it in the 30s). On this particular day I’ve got a cool hat on, and a beige suit. I’m putting hydrogen in my car. Hydrogen increases the speed of combustion, and so it adds to milage — or it has when we’ve added it from electrolysis sources.buxbaum-003

The fun thing about science is that there are always surprises.

Adding hydrogen to a Malibu at REB Research

Adding hydrogen to a Malibu at REB Research

Small hydrogen generators for cooling dynamo generators

2 Replies

A majority of the electricity used in the US comes from rotating dynamos. Power is provided to the dynamos by a turbine or IC engine and the dynamo turns this power into electricity by moving a rotating coil (a rotor) through a non-rotating magnetic field provided by magnets or a non-rotating coil (a stator). While it is easy to cool the magnets or stator, cooling the rotor is challenging as there is no possibility to connect it cooling water or heat transfer paste. One of the more common options is hydrogen gas.

It is common to fill the space between the rotor and the stator with hydrogen gas. Heat transfers from the rotor to the stator or to the walls of the dynamo through the circulating hydrogen. Hydrogen has the lowest density of any gas, and the highest thermal conductivity of any gas. The low density is important because it reduces the power drag (wind drag) on the rotor. The high heat transfer coefficient helps cool the rotor so that it does not burn out at high power draw.

Hydrogen is typically provided to the dynamo by a small hydrogen generator or hydrogen bottle. While we have never sold a hydrogen generator to this market, I strongly believe that our membrane reactor hydrogen generators would be competitive; the cost of hydrogen is lower than that of bottled gas; it is far more convenient and safe; and the hydrogen is purer than from electrolysis.

Purifying the Hydrogen from Browns gas, HHO, etc.

2 Replies

Perhaps the simplest way to make hydrogen is to stick two electrodes into water and to apply electricity. The gas that is produced is mostly hydrogen, and is sometimes suitable for welding or for addition to an automobile engine to increase the mileage. Depending on the electrodes and whether salt is added to the water, the gas that is produced can be Browns gas, HHO,  town gas, or some relative of the three. We are sometimes asked if we can purify the product of this electrolysis, and my answer is typically: “maybe,” or “it depends.”

If the electrode was made of stainless steel and the water contained only KOH or baking soda, the gas that results will be mostly hydrogen and you will be able to purify it somewhat with a polymer membrane if you wish. The gas isn’t very explosive generally, since most of the oxygen that results from the electrolysis will go into rusting out the electrodes. The reaction is thus, H2O + Fe –> H2 + FeO. To see if this is what you’ve got, you can use determine the ratio of gas production with a simple version of the Hoffman apparatus made from (for example) two overturned glass jars, or by separating the electrodes with a paper towel. You can also determine the H2 to O2 ratio (if you know a bit more physics) from a measure of the amperage and the rate of gas production. The hydrogen you form with steel plates will always contain some oxygen though, as well as some nitrogen and water vapor. While a polymer membrane will remove most of the oxygen and nitrogen in this gas, it won’t remove all, and it will not generally remove any of the water. With this gas, I suspect that you would be better off just using it as it is. This is particularly so if the fraction of oxygen is more than a few percent: hydrogen with more oxygen than this becomes quite explosive.

Since this gas will contain water, you probably don’t want to store it, and you probably don’t want to purify it over a metal, either, There are two reasons for this: the water can condense out during storage, and will tend to rust whatever metal it contacts (it’s often alkaline). What’s more, the small amount of oxygen in the hydrogen is likely to react over a hydrogen storage metal to form water and heat. This may give rise to the explosion you were trying to avoid. This is clearly the quick a dirty approach to making hydrogen.

Another version of electrolysis gas, one that’s even quicker and dirtier than the above involves the use of table salt instead of KOH or baking soda. The hydrogen that results will contain chlorine as an impurity, and will be quite toxic, but it will be somewhat less explosive.The hydrogen will smell like bleach and the water you use will turn slightly greenish and quite alkaline. Both the liquid and gas are definitely bad news unless your aim was to make chlorine and alkali; this is called the chlor-alkali process for a reason. On a personal note, as a 12 year old I tried this and was confused about why I got equal volumes of gas on the cathode and anode. The reason was that I was making Cl2, and not O2: the chemistry is 2 H2O + 2 NaCl –> H2 + Cl2 + 2 NaOH. I then I used the bromide version reaction to make a nice sample of bromine liquid. That is, I used KBr instead of table salt. Bromine is brown, oily, and only sparingly soluble in water.

Another version of this electrolysis process involves the use of graphite electrodes. If you are lucky, this will give you a mix of CO and hydrogen and not H2 and O2. This mix is a called “town gas.” It’s a very good welding gas since it is not explosive. It is, however, quite toxic. If you begin to get a headache using this gas stop immediately: you’re experiencing CO poisoning. The reaction here is H2O + C –> H2 + CO. CO headaches just get worse and worse until you die. If you are not lucky here you can get HHO instead of town gas, and this is quite explosive: H2O –> H2 + 1/2 O2. The volume ratio will be a key clue as to which you are making; another clue is to put a small volume in a paper bag and light it. If the bag explodes with a terrific bang, you’ve made the wrong gas. Stop!

With all of these gases I would recommend that you add a polymer of paper membrane in the water between the electrodes. Filter paper will work fine for this as will ceramic paper; the classic membrane for this was asbestos. If you keep the two product gas streams separate as soon as they are formed, you’ll avoid most of your explosion-safety issues. Few people take this advice, I’ve found; they think there must be some simpler way. Trust me: this is the classic, safe way to make electrolysis hydrogen.

A balloon filled with pure hydrogen will not ignite. To show you, here is a 2.5 min long video where I poke a lit cigar into a mylar balloon filled with hydrogen from my membrane reactor generators. Note that this hydrogen does not even burn in the balloon because it is oxygen free. As a safety check try this with your hydrogen, but only on a much-smaller scale. Pure hydrogen will not go boom, impure hydrogen will. My advice: keep safe and healthy. You’ll feel better that way, and your heirs will be less inclined to sue me.

In case you are wondering how electrolysis hydrogen can add to the gas mileage, the simple answer is that it increases the combustion speed and the water vapor decreases the parasitic loss due to vacuum. I’ve got some more information on this here. I hope this advice helps with your car project or any other electrolysis option. In my opinion, one should use a membrane in the water to separate the components at formation in all but the smallest experiments and with the smallest amperage sources. Even these should be done only in a well-ventilated room or on a car that is parked outside of the house. Many of the great chemists of the 1800s died doing experiments like these; learn from their mistakes and stay among the living.

Hydrogen Cylinders versus Hydrogen Generators for Gas Chromatography

4 Replies

Hydrogen is an excellent cover gas for furnace brazing and electronic manufacture; it’s used as a carrier gas for gas chromatography or as a flame-detector gas, and it’s a necessity for ammonia production and most fuel cells. If you are working in one of these fields you can buy bottled hydrogen (cylinders) or a hydrogen generator . The main difference is cost. Cylinder hydrogen is typically the choice for small demand applications. A palladium membrane hydrogen purifier is added ( we make these) if high purity is important. Hydrogen generators are more generally used for larger -demand applications. They are more expensive at the start, but provide convenience and long-term savings. The essay below goes through the benefits and drawbacks of each as applies to gas chromatography.

Point of use Cylinder Hydrogen Is Simple and Allows Easy Monitoring and Control. At the smallest laboratories, those with one or two gas chromatographs, you’ll generally find you are best served by a single hydrogen cylinder for each GC, aided by a hydrogen purifier of some sort. This is called “point of use” hydrogen. Each cylinder is typically belted to a wall and used until the cylinder is empty. At that point, the application is stopped (the purifier is often stopped too) and a new cylinder switched in. There is usually a short break- in period where GC results are unreliable, but after one or two runs, everything is as before. The biggest advantage here is simplicity including ease of pressure control and monitoring. You can always check the pressure right by the GC and adjust it as needed. Long term cost is usually higher, though, and you have to stop whenever a cylinder needs switching.

Multi-cylinder Systems or Generators Provide Fewer interruptions. Larger laboratories tend to use multiple hydrogen cylinders with complex switchover systems, or hydrogen generators. Multiple cylinders are racked together and connected to a manifold and a single, larger purifier (we make these too). Tanks are emptied in series so that there is no disruption. When each tank empties, it is switched out in a way that maintains the flow. One problem is that the pressure and flow does not typically stay constant as the cylinders switch and as additional GCs or other processes are brought on line or taken off.

Purity can suffer too, as there is more tubing and more connections in the system. There is thus more room for leaks and degassing. This can be solved by replacing the single large purifier by point-of-use purifiers, installed just prior to the GC or other application.

Cylinder packs come with a safety disadvantage: with so many cylinders, there is a potential for disastrous leaks or mistakes that empty many cylinders at once — too fast to disperse the large amount of hydrogen released. Maintenance becomes an issue too since the manifolds and automatic switches become complicated quickly. Complex systems can require a trained technician to trouble-shoot and maintain; I sometimes do that, and customers don’t seem to mind, but it’s an issue.

Hydrogen generators can be cheaper and you avoid cylinder changes; Hydrogen generators are fed by tap-water or a very large tank of methanol -water. Running out is less of a problem, and adding more water or methanol to the tank does not affect the hydrogen output.

Safety is improved by limiting the output of the generator to the amount the room will vent. A room with 100 ft3 of air and some circulation can generally host a hydrogen generator 2-3 slpm output with no fear of reaching explosive limits. It’s also worthwhile to fit the hydrogen generator with an alarm or safety that shuts down if a leak is detected (we provide these for purifiers too).

Generator Options: Methanol-based hydrogen generators or electrolysis. Both options are are available in outputs from 250 ccm to 50 slpm. For larger-yet output, you’ll probably want an electrolyzer. In general, either generator will pay for itself in the first year if you use the gas, continuously, or nearly so.

In Electrolytic Hydrogen generators Purified water, either purchased separately, or purified on-site is mixed with an electrolyte, generally KOH, and converted to hydrogen and oxygen by the electrolytic reaction H2O –> H2 + ½ O2.  As the hydrogen produced is generally “wet”, containing water vapor, the hydrogen is then purified by use of a desiccant, or by passage through a metal membrane purifier. Desiccants are cheaper, but the gas is at best 99.9% pure, good enough to feed FIDs, but not good enough to be used as a carrier gas, or for chemical production. Over time desiccants wear out; they require constant monitoring and changing as they become filled with water vapor. Often electrolytic hydrogen generators also require the addition of a caustic electrolyte solution as caustic can leak out, or leave by corrosion mechanisms.

In Reformer-based hydrogen generators a methanol-water mix is pumped to about 300 psi and heated to about 350 °C. It is then sent over a catalyst where it is converted to a hydrogen-containing gas-mix by the reaction CH3OH + H2O –> 3H2 + CO2. Pure hydrogen is extracted from the gas mix by passing it through a membrane, either within the reactor (a membrane reactor), or by use of a membrane purifier external to the reactor.

Cost comparisons. Hydrogen in cylinders is fairly expensive if you use gas continuously. In Detroit, where we are, hydrogen costs about $70 each cylinder low low-purity gas, or $200 for high purity gas. Each cylinder contains 135 scf of gas. If you use 1/10 cylinder per day, you will find you’re spending about $7,300 per year on hydrogen gas, with another $1000 spent on cylinder rental and delivery. This is about the cost of a comparable hydrogen generator plus the water or methanol and electricity run it. If you use significantly less hydrogen you save money with cylinders, if you use more there is significant savings with a generator.

Most hydrogen generators have delivery pressure limitations compared to cylinders. Cylinders have no problem supplying hydrogen at 200 psi or greater pressures. By contrast, generators are limited to only the 60-150 psig range only. This pressure limitation is not likely to be a problem, even for GCs that need higher pressure gas or when the generator must be located far from the  instruments, but you have to be aware of the issue when buying the generator. Electrolysis systems that use caustic provide the highest pressures, but they tend to be the most expensive, and least safe as the operate hot and caustic can drip out. Fuel cell generators and reformers provide lower pressure gas (90 psi maximum, typically), but they are safer. In general generators should be located close to the instruments to minimize supply line pressure drop. If necessary it can pay to use cylinders and generators or several generators to provide a range of delivery pressures and a shorter distance between the hydrogen generator and the application.

Click here for the prices of REB Research hydrogen generators. By comparison, I’ve attached prices for electrolysis-based hydrogen generators here (it’s 2007 data; please check the company yourself for current prices). Finally, the price of membrane purifiers is listed here.

Maintenance required for optimal performance. Often electrolytic hydrogen generators require the addition of a caustic electrolyte solution; desiccant purified gas will require the monitoring and changing of desiccant cartridges to remove residual moisture from the hydrogen. Palladium membrane purifiers systems, and reformer systems need replacement thermocouples and heaters every few years. Understanding the required operating and maintenance procedures is an important part of making an informed decision.

Conclusion:

Cylinder hydrogen supplies are the simplest sources for labs but present a safety, cost, and handling concerns, particularly associated with cylinder change-outs. Generators tend to be more up-front expensive than cylinders but offer safety benefits as well as benefits of continuous supply and consistent purity. They are particularly attractive alternative for larger labs where large hydrogen supply can present larger safety risks, and larger operating costs.

R. E. Buxbaum, January 30, 2013, partially updated Apr. 2022.